In Australia we have the Australian Guide to Healthy Eating, which is essentially a pie chart version of MyPlate, but at least uses pictures of actual food to demonstrate the sorts of foods which belong in each food group. The guide is basically a summary of the observational studies which link diet with adequate nourishment, longevity and chronic disease. The serve sizes I explain as a unit of measurement, a way for nutrition scientists and dietitians to 'count' and describe someone's diet, a way to estimate whether someone is receiving adequate nourishment, not a magically perfect portion that a real person needs to eat at any given time. I explain the jaw-dropping scope for measurement error in nutrition science, both pre and post swallow, and the difference between risk and prediction of disease or longevity in a population vs an individual, and that there is no diet which can guarantee any individual longevity or reliable protection from disease.

Interestingly, dietary quality studies where they assess dietary intake against national food guides using one score for 'core' foods and another score for 'junk' foods (please forgive my non HAESy shorthand) find overwhelmingly that it is in fact not the 'junk' food which increases risks but the lower amount of 'core' foods. In other words, if the 'core' foods are consumed in patterns consistent with what is represented in the guide, the consumption of 'a little' or 'a lot' of 'junk' food makes no difference. Kurotani and colleagues have an excellent summary of this literature in their paper: http://www.bmj.com/content/352/bmj.i1209.full (open access yay!) This body of literature is also nice in that it is the food selection guides for the country that are used to assess the population ie it is sensitive to the value of cultural eating patterns within a population.

To distill the literature even further (if people are interested in experimenting with their diet) it seems that eating a variety of fruit and veg does your body good (so experiment with fruit and veg prep ideas which you find delicious, practical and feasible for you), eating a variety of grain based foods is probably helpful for some (let your body be your guide), dairy food is neither helpful or unhelpful (let your body be your guide) and eating charred meat frequently probably isn't a good idea.

Also important to remember is that the recommendation for low fat dairy foods was reverse engineered on the basis that it would keep energy intake down and thus prevent obesity although there is no evidence to support that this is useful for inclusion in population guides or effective.

So in summary, a population, proportion based food guide can be a useful tool (albeit blunt) for probability based nutrition assessment and as a way to deliver nutrition education about adequate nourishment for human bodies (of all shapes and sizes), but when it is used in a prescriptive way it's definitely then not HAES practice but it also massively oversteps its own evidence base.